Structural Equation Modeling (SEM) is one of the most useful tools we have to determine the mediators and moderates in the integrated models. However, when (or in fact especially since in many cases) the relationships between a good number of variables are mutual (due to ambiguity in direction), the determination of the two roles (moderation vs mediation) becomes complex. As shown here in this simplified model (see the attached picture)- originally developed out of an SEM analysis, we may be able to overcome the confusion by seeking help from SNA (social network analysis).

 The nodes will surely be the variables and the beta coefficients can be used to adjust the connection weights. The nodes with the highest “betweenness” centrality appear to play a stronger role as "mediators" (having more control over the flow of causality) whereas the ones with higher “eigenvector” centrality can be argued to play a more genuine moderating role (even if their influence is low). Closeness may indicate a greater integration between the two roles in the case of those variables with higher centrality.   Mediation and Moderation are then treated as interrelated continuums rather than categories. The interpretation of the model then can be further advanced through the discussion of the centrality of each variable in playing such roles and thus the most effective ways of intervention can be speculated.

In the current mainstream macro- and even meso- sociology, intervention is a rather meaningless concept and this perhaps rather explains the lack of interest among sociologists in such arguments. However, with the growing intensification of social upheavals and the return of conscientious social sciences to the scene, it would be a misconception to think that the lack of adequate socio-historical agency exercised by social scientists (as social scientists) should reduce our arguments to almost mere descriptions of sad realities depicted in the relationships between social factors and phenomena. In a transformative scholarship, a research project is a political one and thus is meant to influence change by empowering the progressive agents of change and any knowledge about the centrality of the factors that can be transferred to the agents of collective action would serve the purpose.